Re: ERB votes on error handling

Peter Flynn wrote:
> 
> I don't believe we can expect any significant advances in browser
> error-handling until we see some editors. One of the most important
> reasons for current HTML browsers' need to handle so many errors is the
> low quality of HTML produced by the average homebrew homepager, whether
> or not they are corporate or private.
> 
> The content providers will define the standard of XML in circulation,
> based on the quality of XML produced by their editors. Right now we're back
> at square 1 where we were in 1991, editing largely by hand in character
> editors without parsed assistance (well, mostly). So mistakes which creep
> in now will acquire the status of "things people do which browsers need to
> accept and get on with".
> 
> Or is this unduly pessimistic?

Not really.  It will create a tear in the 
conformance and a tear in the classes of applications.  
You will now have conforming and non-conforming applications and 
it will be interesting to see where the percentages gather.

None of this worries me.  Good programmers will ignore it 
and applications which fail to be tolerant will be considered 
bad applications.  Some people will write DTDs and parse when 
necessary.

The nutty things the ERB has decided to get WFness are 
almost breathtaking.

len